« Cai Guo-Qiang, critic of collectivism | Main | Who's Your City? »
The ongoing dispute over file-sharing papers
Here is a very interesting article in German on the controversy surrounding Stan Liebowitz, Felix Oberholzer-Gee, and Koleman Strumpf and the issue of whether illegal file-sharing has hurt music sales. An English translation is here; Craig Newmark reports as well. The upshot is that Liebowitz has been attacking the Oberholzer-Gee and Strumpf claim that illegal downloads do not much hurt music sales. Liebowitz claims that the original paper, published in the JPE, is incorrect. In English, here is Liebowitz's side of the story. Here are our earlier posts on the Oberholzer-Gee/Strumpf paper. Here are quotations by Koleman Strumpf. Does anyone have a link to a further defense by the paper's authors?
Posted by Tyler Cowen on March 9, 2008 at 04:11 PM in Economics | Permalink
Comments
Even without having their data, Liebowitz has pretty much destroyed the results of the original paper.
More generally, it seems to be a problem that big journals, particularly in economics, and maybe especially the JPE, are obsessed with counterintuitive results. Wasn't there a study which showed that results in top journals were less likely to be replicated?
And it's pretty shoddy that he hasn't heard back about his comment after half a year. Comments aren't refereed, are they?
Posted by: LemmusLemmus at Mar 9, 2008 5:30:00 PM
Is Spain ,importers and producers of cd burners,blank cds and dvds, heve to pay the "canon" . a surcharge that goes to artists and publishers.So even if you buy a cd for making backups of your computer data you have to pay.But its attacks what you see as the main problem
Posted by: jcm at Mar 9, 2008 7:13:19 PM
Who cares whether free trade in music hurts music industry revenue? The question is whether it increases the amount of music available.
Posted by: chris at Mar 9, 2008 8:16:11 PM
"Who cares whether free trade in music hurts music industry revenue?"
You don't mean trade, you mean theft.
Posted by: LemmusLemmus at Mar 9, 2008 8:49:41 PM
I think if he meant theft he would have said theft. Copyright infringement or illegal/unauthorized copying are more accurate terms (unless you are hiding CDs in your jacket. that really would be theft).
Posted by: nelziq at Mar 9, 2008 9:31:17 PM
Although the question of whether you call it "trade" or "theft" might seem to be of practical interest to some, the answer will not have any real consequences because this trade/theft is as impossible to prevent as impure thoughts are. There will be no going back until another killer technology comes along which has room for a middleman. Writing a paper about it either way is a waste of time.
Posted by: James at Mar 9, 2008 9:34:57 PM
Yes, of course. And evicting someone from a rent-controlled apartment is "stealing" their right to rent below market-clearing prices.
Any government intervention in the market (as copyright clearly is) can be described as a "property" right if you squint hard enough. I guess if it mostly benefits rich people, it doesn't count.
Posted by: chris at Mar 9, 2008 9:45:41 PM
I lack any such defense, in either link or personal statement. Being at present employed by a firm who provides empirical analysis sans data-checking (the client may not even see the estimated coefficients, only the predicted values), I understand why the authors cannot get the file-sharing sites to agree to release their data. I just don't care; if your data cannot be seen by everyone and be checked, why are we to believe results derived from it?
Posted by: hamilton at Mar 9, 2008 10:15:55 PM
I have a dataset that required I sign a confidentiality agreement. That they cannot/will not release the data isn't weird by any means. Despite a policy that requires researchers release their data, how can an editor get around that kind of thing?
But, that still doesn't make it any easier to figure out how is right and who is wrong on this.
Posted by: jason voorhees at Mar 9, 2008 10:47:53 PM
Tyler, since this is on hijacking, no other way to get this out but here. I just read the NY times piece on charity, didn't you write on something similar earlier? Is this hijacking?
Posted by: charity at Mar 9, 2008 11:13:21 PM
The reason why there are copyright laws on music is because the marginal cost of production is zero. People will consume a good until the marginal benefit equals the marginal cost. Without copyright laws, the music would have to be given away. What musician would work so hard to just give her work away?
Posted by: Mike Fladlien at Mar 10, 2008 7:11:16 AM
jason-
The question is not whether it is weird, but whether it is right. Even if I trust that you are a smart and capable researcher who would obviously do everything correctly, why should *everyone* make that assumption? How can the default position be that we trust the researcher got the results correct?
I would argue that, unfortunately, a corner solution is the answer: unless the researcher can provide her data and programming, no publication in a peer-reviewed journal. If we cannot double-check, we simply do not know how the result was acquired.
Posted by: hamilton at Mar 10, 2008 8:24:16 AM
Hamilton - I understand. But I think you're missing the point. Presumably, we sign these confidentiality agreements with firms because otherwise we couldn't get the data. There's no incentive to the firm to hand out their data. A confidentiality agreement allows for the data to be released, without harming the firm (presumably they incur some cost if the data was public use).
So either way, you get the same outcome. On the one hand, you get results we cannot replicate because we cannot get the data. On the other hand, we get no results to replicate at all. Personally, I prefer the former to the latter. At least with the former, it's still an option for someone else to approach the same firm and get the data, or to get a similar firm's data but agreeing to the same conditions. But if we cannot make confidences at all, then there's a lot of data we cannot acquire. I think on average, we'll learn more by having these agreements. But this isn't easy. This is just the real world.
Posted by: jason voorhees at Mar 10, 2008 9:24:42 AM
My guess, and it is just a guess, is that music downloading does lower sales and revenues, but increases the net quantity of music consumed and produced, and therefore adds to net utility.
Why? Well, other studies have shown that music downloading may decrease sales for the more popular artists, but they increase sales of more marginal artists. The net effect on sales may be negative, but I believe that the marginal artists are more elastic producers. They need the money to keep making and distributing music.
I might also add that music downloading, by promoting more marginal artists' sales at the expense of popular artists' sales, counterbalances a distortion in the ASCAP system. Under ASCAP, less popular artists do not get paid for their music that is played on the radio, because ASCAP doesn'y pay an artist royalties unless their number of radio plays reaches a certain threshhold.
As a result, ASCAP pays popular artists royalties even for the radio play of less popular artists. ASCAP basically steals from less popular artists and gives their royalty money to more popular artists. Note, the music industry has been quite happy with this arrangement, given that it benefits "their" artists. This property rights violation hasn't caught the industry's attention, for some mysterious reason.
So, it certainly looks to me that society has no real interest in expending valuable resources to stop music downloading, even when it violates copyright.
Posted by: Keith at Mar 10, 2008 11:10:58 AM
nelzig,
"I think if he meant theft he would have said theft. Copyright infringement or illegal/unauthorized copying are more accurate terms"
Formally-legally you're right, of course. Consider "theft" a methaphor. It's certainly a better one than "trade".
chris,
"Yes, of course. And evicting someone from a rent-controlled apartment is "stealing" their right to rent below market-clearing prices.
If they paid the rent, you could argue this - regardless of what you think about rent control. Otherwise, I don't see the parallel.
"Any government intervention in the market (as copyright clearly is)"
Not really. Any artist is allowed to give away his stuff for free.
"I guess if it mostly benefits rich people, it doesn't count."
Most professional musicians are poor.
Posted by: LemmusLemmus at Mar 10, 2008 12:38:58 PM
Consider "theft" a methaphor. It's certainly a better one than "trade".
It may be a better metaphor than trade, but it's still not a good one. Copying music is significantly different than normal classes of theft in that there is no direct loss to anyone. People who intentionally obscure this point are annoying. It is an important point b/c it leads to the observation that perhaps this file sharing is a net social benefit. If so, it suggests that we not pass new laws to make it more difficult, and, in fact, we should not get so worked up about it. In a similar way, if we decided that rent-controlled apartments were not a great social benefit, we might make it easier, legally, for landlords to evict their underpaying rentors.
Posted by: mpowell at Mar 10, 2008 1:12:48 PM
"Most professional musicians are poor."
Yes, but music downloading increases the sales for these poor musicians. The sales losses are felt among the top musicians. Downloading increases sales for the bottom 75% of all musicians.
http://www.katallaxi.se/grejer/blackburn/blackburn_fs.pdf
Check out Table 8 on page 46.
Posted by: Keith at Mar 10, 2008 1:14:37 PM
